The UCLA project develops and applies structural and bioinformatic tools for the discovery of protein drug targets and drugs against tuberculosis. We have developed the database for the TB Structural Genomics Consortium. This contains a wealth of information on the genome and proteome of Mycobacterium tuberculosis, which is open to researchers world wide to aid them in their structural studies of MTb proteins. Improvements to the database will enhance the ease with which scientists can view and update information about each protein. The information will include the complete history of experiments on each protein, as the project moves from cloning and expression, through purification, crystallization, structure determination, and drug discovery. The database will also incorporate new bioinformatics methods being developed at UCLA. These include methods for inferring the biological function of proteins from their structures by the server ProKnow and for inferring biological function from genome sequences from the database ProLinks. A new method for inferring biological function from combined genome sequences and microarray data will be developed to aid consortium members in selecting the most effective targets for structural studies in the effort to determine drugs. Structural studies of Mtb proteins at UCLA will focus on secreted, regulatory, and metabolic proteins. These proteins are involved in the invasion of host cells by Mtb, in the protection of Mtb against host cell defenses, in survival of Mtb and are thus potential protein drug targets. Secreted proteins of Mtb in particular are excellent anti- Mtb drug targets because drugs inhibiting these proteins do not have to penetrate the waxy cell wall of Mtb. Structures of each of these protein targets can then be screened in silico against ligands, in collaboration with Dr. James Sacchettini's laboratory. Predicted ligands will be co-crystallized for further structural studies. Potential inhibitors will be studied in collaboration with Dr. William Jacobs'laboratory to see if they prevent growth of Mtb within a mouse model. Each protein whose structure we determine, we will also characterize biochemically. Developing an in vitro assay will allow high-throughput in vitro drug screening, continuing our collaboration with the Southern Research Institute. All lead compounds will be cocrystallized with its target protein for structural studies. Where functions of proteins are unknown, our bioinformatics tools may predict a function, which will then be verified experimentally, to move the protein into the drug discovery pipeline. Dozens of MTb proteins are in the UCLA pipeline, with some half dozen at the stage of crystals.